Вы находитесь на странице: 1из 10

Santosh Krishnamurthy

Yoav Peles
Department of Mechanical, Aerospace, and
Nuclear Engineering,
Rensselaer Polytechnic Institute,
Troy, NY 12180

Flow Boiling Heat Transfer on


Micro Pin Fins Entrenched in a
Microchannel
Flow boiling of 1-methoxyheptafluoropropane (HFE 7000) in 222 m hydraulic diameter channels containing a single row of 24 inline 100 m pin fins was studied for mass
fluxes from 350 kg/ m2 s to 827 kg/ m2 s and wall heat fluxes from 10 W / cm2 to
110 W / cm2. Flow visualization revealed the existence of isolated bubbles, bubbles interacting, multiple flow, and annular flow. The observed flow patterns were mapped as a
function of the boiling number and the normalized axial distance. The local heat transfer
coefficient during subcooled boiling was measured and found to be considerably higher
than the corresponding single-phase flow. Furthermore, a thermal performance evaluation comparison with a plain microchannel revealed that the presence of pin fins considerably enhanced the heat transfer coefficient. DOI: 10.1115/1.4000878
Keywords: micro pin fins, heat transfer coefficient, flow maps

Introduction

The level of device integration and the clock speed of microelectronic devices have been steadily increasing over the last several decades. This has resulted in a continuous increase in the
power density of electronic devices, giving rise to the need for
aggressive and effective cooling solutions. The recent development in microfabrication technology has enabled a new class of
heat sinks for high heat flux applications made of microelectromechanical system MEMS based microchannel. Microchannels
with micro pin fins have been receiving attention because they can
significantly enhance the performance of heat sinks compared
with plain microchannels. As a result, a research on flow boiling
and single-phase flow in micro-pin fin heat sinks has been carried
out by several groups 112. It is well known that heat transfer
during subcooled flow boiling can be significantly enhanced compared with single-phase liquid flow. However, very limited studies
have been performed to elucidate the heat transfer characteristics
of subcooled boiling in micro pin fin configurations.
In conventional scale, subcooled flow boiling has been extensively studied for in-tube systems and various models have been
developed to predict the heat transfer rates. These methods can be
broadly classified into three categories, namely: empirical correlations to predict the wall heat flux 1317, empirical correlations
to predict the partitioning of wall heat flux 14,18,19, and mechanistic correlations for the total heat flux and their partitioning
2022. Warrier and Dhir 23 provided a detailed review on
these various methods and concluded that mechanistic correlations give a better physical insight into subcooled boiling phenomenon and proposed the use of submodels to define the bubble
dynamics. Literature review reveals that boiling in crossflow systems has been extensively studied, but very limited studies have
been reported in subcooled boiling regime for these configurations. Huang and Witte 24,25 studied the effect of liquid flow
and high subcooling across a bank of horizontal tube bundles and
developed a correlation for the heat transfer coefficient using the
methodology by Chen 26 developed for saturated nucleate boiling. Shah 27 compared his correlation for heat transfer coefficient around a single-tube, developed for highly subcooled liquid,
Contributed by the Heat Transfer Division of ASME for publication in the JOURHEAT TRANSFER. Manuscript received November 24, 2008; final manuscript
received September 30, 2009; published online February 19, 2010. Assoc. Editor:
Satish G. Kandlikar.

NAL OF

Journal of Heat Transfer

with various correlations developed for low subcooled liquid and


found good agreement. Cornwell 28 studied saturated nucleate
flow boiling in tube bundles and attributed the enhancement in
heat transfer rates to sliding bubbles. Gupta 29 attributed the
enhancement in the heat transfer coefficient during nucleate boiling to the turbulence caused by the presence of bubbles on the
surface.
It has been shown that reduction in length scale has a considerable affect on the bubble dynamics and flow characteristics during flow boiling. For subcooled flow boiling this is especially
important since the heat transfer mechanism is very dependent on
the bubble nucleation process. For example, Lee et al. 30 and
Kuo et al. 31 examined the bubble dynamics for water in a
microchannel and observed considerable differences compared
with conventional scale systems. Cognata et al. 5 performed a
visual study on flow pattern in staggered square micro pin fins and
observed bubbly, slug, and annular flows. Krishnamurthy and Peles 3 investigated flow boiling heat transfer of water across
densely packed staggered micro pin finsprimarily focusing on
saturated flow boilingand found that convective boiling was the
dominant heat transfer mechanism.
Based on previous studies, it can be concluded that knowledge
about bubble characteristics in diminishing length scales is important to elucidate heat transfer mechanisms. In the current manuscript, flow boiling across a single row of inline micro pin fins
entrenched in a microchannel is presented. The microdevice consists of five 200 m wide and 243 m deep microchannels, each
equipped with an inlet orifice, consisting of 24 columns of
100 m diameter circular pin fins, with pitch-to-diameter ratio of
4. This study aims to elucidate the local heat transfer coefficient,
to identify flow patterns, and to decipher the heat transfer mechanisms. Additionally, the performance of the pin fin device was
compared with a plain microchannel device.

Device Design

A computer aided design CAD schematic of the microdevice,


consisting of five 200 m wide and 243 m deep microchannels
entrenched in a 1800 m wide channel, is shown in Fig. 1. Each
microchannel encompassed 24 inline 100 m diameter micro pin
fins with a pitch-to-diameter ratio of 4. Pressure taps were placed
at the inlet and the exit of the microchannel array to enable pressure measurements. A micro-orifice, 400 m long and 20 m
wide, was fabricated upstream of each microchannel to suppress

Copyright 2010 by ASME

APRIL 2010, Vol. 132 / 041007-1

Downloaded From: http://asmedigitalcollection.asme.org/ on 04/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 1 Device overview showing the device dimensions

flow oscillations. A heater was deposited on the backside of the


channel-pin fin section excluding the orifice to provide the requisite heat flux. A thermistor 10 m wide and 300 m long was
placed 3.33 mm from the channel inlet. A Pyrex cover sealed the
device from the top and allowed flow visualization. For details
regarding the experimental set up and microfabrication process
flow, please refer to Ref. 3.

x=

c pTsat Ti
P QlossLx/Lo m
h fg
m

The local heat transfer coefficient for the microchannel with pin
fins was calculated according to

Data Reduction

The voltage and current were used to calculate the input power,
while the local temperature from the thermistor was obtained from
the calibration curve. Assuming 1D steady state conduction
through the silicon block, the local surface temperature of the
device was obtained by
Tx,s = Tthermistor

P Qlossts
k sA p

The local quality was calculated from the known mass flow rate
according to

Table 1 Uncertainty of variables


Uncertainty variable

Measurement range

Error

018 ml/min
040 V
05 A
NA
NA
NA
NA
NA
NA
01379 kPa
NA
NA
NA
NA
NA
NA
NA

2%
0.5%
0.5%
1 C
1%
0.67%
0.5%
3.4%
0.5 C
3.5 kPa
3%
1216%
1216%
1015%
1015%
3.4%
11%

Flow rate, Q a
Voltage, V
Current, I
Ambient temperature, Tamb
Channel width, w
Channel height, H
Density of liquid, l
Mass flux, G
Average surface temperature, Tr
Pressure, p
Heat flux, qw
Local thermal resistance Rconv
Local heat transfer coefficient, hx
Average heat transfer coefficient, h
Average Nusselt number, Nu
Reynolds number, ReD
Boiling number, Bo
a

The manufacturer provided the flow rate range for water, while experiments were
performed with HFE-7000. The flow meter was calibrated prior to experiments.

041007-2 / Vol. 132, APRIL 2010

Fig. 2 Images showing: a isolated bubble region I, b


bubble interacting BI, c multiple flow region M, and d
annular flow A

Transactions of the ASME

Downloaded From: http://asmedigitalcollection.asme.org/ on 04/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 3 Flow maps based on wall heat flux for all the mass fluxes: a G = 350 kg/ m2 s, b G = 564 kg/ m2 s, c G
= 689 kg/ m2 s, and d G = 827 kg/ m2 s

hx =

P Qloss
AtTx,s T1

where T1 = Tmx if x 0 and T1 = Tsat if x 0. The mean local fluid


temperature Tmx was obtained through an energy balance

P Qloss Lx
Tmx = Tmi +
Cp L
m

The total surface area of the channel is given by


At = N f f As,f + Ab + N p pAs,p
where

At,plain = N f f As,f + At N f As,f

The convective resistance used to evaluate the thermal performance is


Rconv =

1
h xA t

The statistical average value of the heat transfer coefficients hx


and the corresponding Nusselt numbers over all heat fluxes for a
fixed Reynolds number in the single-phase region was obtained by
M

f =

tanhm f H
,
mfH

tanhm pH
p =
,
m pH

mf =

mf =

hx2W + L
,
ksWL

hxDH
D2
ks
4

The above equations Eqs. 3 and 5 were solved iteratively


to obtain the local heat transfer coefficient. Similar methodology
was adopted to obtain the heat transfer coefficient for the plain
channel, but the total surface area was calculated as
Journal of Heat Transfer

h = 1
hx,i
x
M i=1

Nux =

1
Nux,i
M i=1

The uncertainties associated with the measured values were obtained from the manufacturers specification sheets Table 1
while the uncertainties associated with the derived quantities were
obtained by using the propagation of uncertainty analysis, and are
also given in Table 1.
APRIL 2010, Vol. 132 / 041007-3

Downloaded From: http://asmedigitalcollection.asme.org/ on 04/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 4 Flow map showing the different flow patterns along the
channel

Results and Discussion

4.1 Flow Pattern. In order to determine the dominant flow


patterns in the device, images were taken at 10 different locations
along the length of the channel. The flow patterns were then
manually classified into bubbly flow, multiple flow M, and annular flow A. The bubbly flow was further categorized into two
regions: isolated bubbles I and bubbles interacting BI. The
isolated bubble region extends over a relatively small section of
the channel, where bubbles nucleated and departed without coalescing with bubbles from neighboring sites Fig. 2a. With increasing nucleation site density, bubbles began to coalesce Fig.
2b, and formed larger bubbles with diameter smaller than the
pin in diameter. Further downstream, these bubbles grew and developed into vapor slugs, which were intermittently sheared by the
pin fins and broken into bubbles again. This region where bubbles
and vapor slugs coexisted was termed multiple flow Fig. 2c.
Eventually, the vapor slugs merged and the flow transitioned to
annular flow, where liquid traversed through the channel walls,
while vapor propagated through the corenucleation was suppressed Fig. 2d. In the channel inlet, only liquid single-phase
was present. As expected, for a given mass flux, all flow patterns
shifted upstream with increasing heat flux. Figures 3a3d
show a flow maps for all mass fluxes as a function of wall heat
flux and normalized axial distance. The hatch regions on the flow
map indicate flow transition regions. Visualization measurements
were repeated three times to obtain a meaningful statistical average transition zone. In order to obtain a more general flow map
using all mass fluxes and heat fluxes, an attempt to collapse the
four flow maps into a single map using the boiling number was
carried out. This was done by first plotting the above flow maps in
terms of the boiling number and normalized axial distance, followed by fitting the data points with a best curve fit, as shown in
Fig. 4. 90% of the transition data points fell within 12% of the
transition lines, as shown by the dotted lines in the figure.
4.2 Heat Transfer Coefficient. The local single-phase heat
transfer coefficient is shown as a function of the wall heat flux for
different mass fluxes in Fig. 5a. The heat transfer coefficient
characteristics followed similar trend to those observed in conventional scale tube bundle systems, i.e., independent of wall heat
flux and increasing with mass fluxes. Figure 5b also shows the
variation in the Nusselt number Nuhdefined based on channel
hydraulic diameteras a function of the Reynolds number ReD.
The local heat transfer coefficients during boiling as a function of
wall heat flux are shown for different mass fluxes in Figs.
041007-4 / Vol. 132, APRIL 2010

Fig. 5 a The variation in single-phase heat transfer coefficient as a function of wall heat flux and b variation in Nusselt
number as a function of the Reynolds number

6a6d the exit pressure was maintained at 260 kPa. For G


= 350 kg/ m2 s, the heat transfer coefficient increased linearly
with wall heat flux for qw 45 W / cm2, and subsequently decreased. Similar decrease in the heat transfer coefficient was also
observed for G = 564 kg/ m2 s at qw = 80 W / cm2. For mass fluxes
of 689 kg/ m2 s and 827 kg/ m2 s, the heat transfer coefficient
increased linearly with wall heat flux. Figure 7 shows the variation in the heat transfer coefficient as a function of the local quality for the four mass fluxes. The majority of the datum points
corresponding to two-phase flow are in the subcooled boiling regime. Figure 7 also shows that the heat transfer coefficient during
subcooled boiling is considerably higher compared with the
single-phase heat transfer coefficient. Such an enhancement during heat transfer coefficient in nucleate boiling has been observed
in both conventional scale channels and minichannels 32,33 and
has been attributed to various mechanisms, such as the evaporation of the microlayer beneath a growing bubble, transient conduction through the cold liquid layer replacing the superheated
liquid layer carried away by the departing bubble 21, and sliding
bubbles 28. The contribution of these mechanisms are added
linearly to account for the total heat flux according to 21

qt = qev + qtr + qsp

10

where qt, qev, qtr , and qsp


are the total heat flux, evaporative heat
flux, quenching heat flux, and single-phase heat flux, respectively.
Transactions of the ASME

Downloaded From: http://asmedigitalcollection.asme.org/ on 04/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 6 Heat transfer coefficients as a function of wall heat flux: a G = 350 kg/ m2 s, b G = 564 kg/ m2 s, c G
= 689 kg/ m2 s, and d G = 827 kg/ m2 s

The contribution of the above mentioned mechanisms toward the


observed enhancement in the heat transfer is assessed below.

Heat Transfer Enhancement Mechanisms

Evaporative heat flux. The evaporative heat flux is defined as


the latent energy carried away by the bubbles per unit area and

consists of two terms arising from the stationary and sliding


bubbles

+ qev,sl
= fVdvNah fg + fVl VdvNah fg
qev = qev,st

The stationary bubble heat flux is associated with the growth of


the bubble nucleating on the surface, and the sliding bubble is
associated with the heat transfer during bubble growth while moving along the surface. In Eq. 11, the bubble departure frequency
was assumed to be equal to the bubble lift off frequencyan
assumption that is based on observation of many bubbles. Assuming the departing bubble to be spherical, the volume of the departing and lifting bubbles can be calculated as

Vd =

Fig. 7 The variation in heat transfer coefficient as a function of


local quality heat flux in W / cm2 in parenthesis

Journal of Heat Transfer

11

D3d
,
6

Vl =

Dl3
6

12

Quenching heat flux. As a bubble departs from a nucleation site,


it displaces superheated liquid adjacent to the wall by cold liquid
from the bulk flow. Han and Griffith 34 postulated that the departing bubble carries away with it liquid from an areatermed
area of influencethat is four times the projected area of the
departing bubble. The quenching heat flux was obtained by following the approach adopted by Mikic and Rohsenow 35 assuming pure conduction through the liquid in the area of influence.
For any stationary bubble departing from the site, the total average heat flux over the area of influence is given by
APRIL 2010, Vol. 132 / 041007-5

Downloaded From: http://asmedigitalcollection.asme.org/ on 04/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Table 2 Bubble dynamics in the isolated bubble region z = 2.5 3.5 mm


G

qch

Nwall

N=0

N=180

f wall

f =0

f =180

Dd,wall

Dd,=180

Dd,=0

qev
%

qtr
%

350
350
565
565
689
689
827
827

13
16.05
19.2
20.3
27.1
29.9
36.2
39.5

2
5
5
10
16
20
20
25

2
2
2
3
3
6
4
5

3
4
3
5
4
4
7
12

3511
4300
4712
3472
2550
2819
2260
3805

413
1592
2799
3235
4749
5095
5885
6297

50
445
2175
2233
481
468
250
110

32
34
32.5
40.3
37.7
33.6
41.9
38

18
14
18.5
13
10.02
10.02
9.22
8.9

72
55
74.6
70.6
43
30
42
25

0.62
1.77
2.6
4.8
2.5
2.3
2.8
3.96

8.12
7.38
8.9
12.1
5.2
5.9
7.5
8.9

= 2kc pl fD2dNaTw Tl
qtr,st

13

In addition to the stationary bubbles, the bubbles sliding along the


surface also displaced the liquid from the surface. Assuming transient conduction through the displaced liquid layer, the total average heat flux over the area swept by the bubble is given by

14

kcpl fAslNaTw Tl

15

=f
qtr,st

kTw Tl

AslNa dt

The sliding area for the bubbles was obtained as follows:


Asl = Davl =

Dd + Dl
l
2

16

l is the sliding distance, which was obtained through flow visualization. The bubble lift off diameter was found to be approximately four times the bubble departure diameter for bubbles
nucleating from the frontal stagnation point = 0 deg and twice
the bubble departure diameter for bubbles departing from the sidewalls.
It should be noted that the use of Han and Griffiths quenching
term might not be entirely adequate to sliding bubbles since a
sliding bubble is merely moving the superheat liquid layer along
the wall rather than displacing it from the surface. However, it can
be argued that because of this the model should overpredict the
contribution of a sliding bubble. The analysis about the heat transfer mechanisms discussed in the paper will later show that the
contribution of the quenching to the total heat flux is insignificant
and therefore, that quenching is not an important heat transfer
mechanism in this study. In other words, the use of the quenching
model by Han and Griffith 34 serves in this study to demonstrate
that it is not an important mechanism rather than to obtain accurate measure of the quenching effect.
The total heat transfer through transient conduction can be
added linearly as

+ qtr,sl

qtr = qtr,st

17

In order to determine the above heat fluxes, the bubble dynamics


on the heated wall, such as bubble departure frequency, nucleation
site density, and bubble departure diameter are necessary, and
were obtained through flow visualization technique discussed in
Ref. 36. Since clear bubble images were required to obtain the
parameters defining the bubble dynamics, all images were taken in
the isolated bubble region. Figures 6a6d also show the conditions under which the isolated bubbles were observed shown by
arrows. Table 2 gives values of various parameters defining the
bubble dynamics between z = 2.5 mm and z = 3.5 mm. Table 2
also shows the contribution of the evaporative and transient conduction heat fluxes to the total heat flux. Since the bubbles emanated from two angular positions = 0 deg, = 180 deg 36,
the bubble dynamics parameters for these locations were included
in the calculation of the evaporative and transient heat fluxes. The
041007-6 / Vol. 132, APRIL 2010

contributions of both the evaporative and the transient conduction


heat fluxes toward the total heat flux were insignificant. This
shows that another significant mechanism in the form of bulk fluid
agitation is a more potent heat transfer mode than the local heat
removal or just the motion of the superheated layer adjacent to
the wall. Basu et al. 21 stated that for regions between the onset
of nucleate boiling and onset of significant voidthe region
where the bubble interaction begin to dominatethe enhancement
in the heat transfer coefficient was mainly due to single-phase
convection, which was enhanced by up to 30% as a result of the
presence of bubbles on the surface. The enhancement in the heat
transfer coefficient in the current study ranges between 5090% in
the isolated bubble region, which was larger than those observed
in conventional scale systems. Based on the values obtained for
the evaporative and transient conduction heat fluxes in the current
study, it can be concluded that in the isolated bubble region, the
observed enhancement in the heat transfer coefficient is neither
due to microlayer evaporation nor due to the transient conduction
through the liquid layer. It appears that the presence of bubble in
the flow has a more pronounced influence on the heat transfer
characteristics in the current microscale study than in large scale
systems. The relatively large bubble diameter-to-channel hydraulic diameter ratio, compared with conventional scale systems, can
significantly alter the flow characteristics in the channel, and thus,
the heat transfer mechanisms. For example, the Reynolds number
calculated for a bubble of largest diameter approximately
75 m and the highest mean flow velocity in this study
827 kg/ m2 s is 124, which corresponds to laminar flow. It can
thus be postulated that the bubbles growing on the sidewalls and
the pin fins induce wakes downstream the channel very much
similar to the vortex shedding observed in classical fluid dynamics
such as flow across cylinders and spheres, which disrupts the
boundary layer significantly beyond the region immediately adjacent to the bubble resulting in higher heat transfer coefficient. The
small length scale of the channel can significantly amplify the
bubble agitation effect. Additionally, the presence of recirculation
zone upstream and downstream the pin fin can also contribute
significantly toward enhanced mixing, and thus, enhanced heat
transfer. Therefore, at low qualities, the observed enhancement
can be attributed to bubble agitation and perturbation of the
boundary layer. Similar argument was recently made by Donnelly
et al. 37 who studied flow across a sliding bubble on an inclined
surface and concluded that bubble induced wakes contribute significantly toward heat transfer enhancement. At high qualities, the
nucleation site density increased and bubbles began to merge.
This in turn increased the evaporation and transient conduction
contributions to the total heat flux. At higher qualities multiple
bubble interaction region, the heat transfer enhancement might be
due to similar mechanisms observed in a conventional scale.
5.1 Comparison With a Plain Microchannel. To study the
effect of the pin fins on the heat transfer coefficient, experiments
with plain microchannels were also conducted under similar thermal hydraulic conditions and the thermal resistances of the two
devices were compared. The total thermal resistance consists of
Transactions of the ASME

Downloaded From: http://asmedigitalcollection.asme.org/ on 04/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 8 Comparison of convective resistance as a function of


mass flow rate for both devices

Fig. 9 Enhancement in the single-phase Nusselt number for


different Reynolds numbers

conductive resistance Rcond resulting from heat conduction from


the base of the silicon block to the channel surface; sensible heat
resistance due to heating of the liquid Rheat; and convective resistance Rconv resulting from convection from the channel walls

to the fluid. Since the experiments were conducted at similar mass


flow rates for both devices, the resistances due to sensible heating
were the same. Likewise, the conductive resistance was the same
for the two devices. Thus, only the convective resistance was

Fig. 10 Comparison of heat transfer coefficients for a microchannel with pin fins and a plain microchannel for different
mass fluxes: a Gch = 282 kg/ m2 s, b Gch = 345 kg/ m2 s, and c Gch = 413 kg/ m2 s

Journal of Heat Transfer

APRIL 2010, Vol. 132 / 041007-7

Downloaded From: http://asmedigitalcollection.asme.org/ on 04/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 11 Comparison of thermal convective resistances for both devices: a Gch = 282 kg/ m2 s, b Gch = 345 kg/ m2 s, and
c Gch = 413 kg/ m2 s

evaluated and compared. Figure 8 compares the single-phase convective resistance Eq. 7 as a function of mass flow rate for
both microchannel systems. The convective resistance decreases
with increasing mass flow rate for both devices. The lower thermal resistance of the microchannels with pin fins compared with
the plain microchannel is a result of the higher heat transfer coefficient and larger surface area. Figure 9 compares the ratio of the
Nusselt numberbased on the channel hydraulic diameterfor
both devices as a function of Reynolds number ReD defined
based on hydraulic diameter of the channel. The enhancement of
the heat transfer coefficient increased from 1.3 to 3 when the
Reynolds number increased from 84 to 197. Therefore, in addition
to the surface area enhancement Apin = 1.25 Aplain, the presence
of pin fins significantly affects the hydrodynamic characteristics
of the flow resulting in increased heat transfer coefficient. The
enhancement increase with the Reynolds number can be attributed
to the wake interaction between the pin fins. At low Reynolds
number, the wake interaction is less rigorous, and thus, a lower
enhancement in the heat transfer coefficient was observed. But at
higher Reynolds number, the interaction between the wakes increased, promoting advection mixing, and thus, reducing the
thermal resistance. Figures 10a10c show the heat transfer coefficient as a function of mass quality for both devices. The heat
transfer coefficient followed similar trend with respect to quality,
but was quantitatively lower for the plain microchannel for all
mass fluxes. The comparison of the convective resistance as a
041007-8 / Vol. 132, APRIL 2010

function of local quality Figs. 11a11c also shows that the


resistance is lower for the microchannel with pin fins. It is also
evident that the rate at which the convective resistance decreases
for the plain microchannel is higher compared with that observed
in the microchannel with pin fins. This can partly be attributed to
the more rapid decrease in the fin efficiency of the micro pin fins
compared with that of channel sidewalls. Thus, with increasing
heat flux, the effective surface area of the microchannel with pin
fins decreased more rapidly compared with that of plain microchannel. As a result, the resistance of the microchannel with pin
fins decreases more moderately. Nevertheless, the heat transfer is
still enhanced by convective mixing of the pin fins, which lowers
the convective resistances compared with plain microchannel. The
enhancement in heat transfer coefficient is quantified by an enhancement factor E p defined as
Ep =

hpin fin
hplain

18

Figure 12 shows the variation in the enhancement factor for different qualities. The enhancement in the two-phase heat transfer
coefficient is smaller when compared with the enhancement observed during single-phase. As discussed previously, the enhancement in the two-phase heat transfer coefficient in the isolated
bubble region is due to bubble agitation and the perturbation of
the boundary layer by the bubbles. Unlike in channel with microTransactions of the ASME

Downloaded From: http://asmedigitalcollection.asme.org/ on 04/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

for Multi-Phase, Nanotechnology-Enhanced Cooling of High


Power Microelectronic Systems. The microfabrication was performed in part at the Cornell NanoScale Facility a member of the
National Nanotechnology Infrastructure Network, which is supported by the National Science Foundation under Grant No. ECS0335765, its users, Cornell University, and industrial affiliates.

Nomenclature
Ap
At
Asl
Bo
cp
D
Dh
Dd
Dl
Ep
f
G
h fg
hx
H
I
ks
L

m
Na
Nf
Np
Nux

Fig. 12 Enhancement in the heat transfer coefficient for a microchannel with pin fins as a function of mass quality

pin fins, where the presence of pin fins shadows the agitation
affect, in plain microchannel, such agitation can affect extended
regions of the channel. As a result, the heat transfer coefficient for
the plain channel increases at a higher rate compared with microchannel with pin fins in the isolated bubble region. But with increasing heat flux, for the microchannel with pin fins in the multiple bubble interaction region, the heat transfer coefficient
increases significantly due to convective mixing aided by the presence of the pin fins. It follows that the enhancement increases
after reaching a minimum, which was observed for both Gch
= 282 kg/ m2 s and Gch = 417 kg/ m2 s.

Summary

Flow visualization revealed the existence of isolated


bubbles, bubbles interacting, multiple flow pattern, and annular regions along the channel length. The observed flow
patterns were mapped as a function of the boiling number
along the channel length.
Single-phase heat transfer coefficient for the microchannels
with pin fins was found to be considerably higher compared
with the plain wall channels. This was attributed to a combination of enhanced area and mixing.
Considerable enhancement in the heat transfer coefficient
during subcooled boiling over the corresponding singlephase heat transfer coefficient was observed. In the isolated
bubbles region, this enhancement was attributed to the agitation of the liquid due to bubble protrusion and disruption
of the boundary layer.
The heat transfer coefficient during subcooled boiling for
the microchannel with pin fins was higher than the corresponding value for plain microchannel. But the enhancement in the heat transfer coefficient was smaller in comparison to that observed during single-phase flow, especially in
the isolated bubble region. This was attributed to the reduction in fin efficiency.

Acknowledgment
This work is supported by the Office of Naval Research ONR
under the Multidisciplinary University Research Initiative
MURI Award No. GG10919 entitled System-Level Approach
Journal of Heat Transfer

platform area m2
total surface area m2
sliding area swept by bubbles m2
boiling number qch / Gh fg
specific heat capacity kJ/ kg K
diameter of pin fin m
hydraulic channel diameter m
bubble departure diameter m
bubble lift off departure diameter m
enhancement factor
bubble departure frequency Hz
mass flux kg/ m2 s
latent heat of vaporization kJ/kg
local heat transfer coefficient W / m2 K
height of microchannel m
current A
substrate thermal conductivity W / m K
length of the channel m
mass flow rate kg/s
nucleation site density
number of fins
number of pin fins
local Nusselt number based on characteristic
length scale hD / k f ; hDh / k f
power W
evaporative heat flux W / m2
stationary bubble evaporative heat flux
W / m2
sliding bubble evaporative heat flux W / m2
stationary bubble quenching heat flux W / m2
sliding bubble quenching heat flux W / m2
heat loss W
Reynolds number based on channel hydraulic
diameter GDh /
thermal convective resistance K/W
thermal conductive resistance K/W
resistance due to sensible heating of fluid
K/W
liquid temperature K
local fluid temperature K
inlet fluid temperature K
saturation temperature K
local surface temperature K
temperature of thermistor K
substrate thickness m
voltage V
width of rectangular pin fin m
vapor quality

fin efficiency
pin fin efficiency
vapor density
radial angle

P
qev
qev,st

Subcooled and low quality saturated flow boiling across micro


pin fins entrenched in a microchannel was studied for various
mass fluxes and heat fluxes. The following summarizes the main
findings of this study.

qev,sl

qtr,st

qtr,sl

Qloss
ReD

Rconv
Rcond
Rheat
Tl
Tmx
Tmi
Tsat
Tx,s
Tthermistor
ts
V
W
x
Symbol

f
p
v

References
1 Qu, W., and Siu-Ho, A., 2008, Liquid Single-Phase in an Array of Micro-PinFinPart I: Heat Transfer Characteristics, ASME J. Heat Transfer, 13012,

APRIL 2010, Vol. 132 / 041007-9

Downloaded From: http://asmedigitalcollection.asme.org/ on 04/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

p. 122402.
2 Qu, W., and Siu-Ho, A., 2008, Liquid Single-Phase in an Array of Micro-PinFinPart II: Pressure Drop Characteristics, ASME J. Heat Transfer,
13012, p. 124501.
3 Krishnamurthy, S., and Peles, Y., 2008, Flow Boiling of Water in a Circular
Staggered Micro-Pin Fin Heat Sink, Int. J. Heat Mass Transfer, 5156, pp.
13491364.
4 Siu-Ho, A., Qu, W., and Pfefferkorn, F., 2007, Experimental Study of Pressure Drop and Heat Transfer in a Single-Phase Micropin-Fin Heat Sink,
ASME J. Electron. Packag., 1294, pp. 479487.
5 Cognata, T. J., Hollingsworth, K. D., and Witte, L. C., 2007, High-Speed
Visualization of Two-Phase Flow in a Micro-Scale Pin-Fin Heat Exchanger,
Heat Transfer Eng., 2810, pp. 861869.
6 Prasher, R. S., Dirner, J., Chang, J.-Y., Myers, A., Chau, D., He, D., and Prstic,
S., 2007, Nusselt Number and Friction Factor of Staggered Arrays of Low
Aspect Ratio Micropin-Fins Under Cross Flow for Water as Fluid, ASME J.
Heat Transfer, 1292, pp. 141153.
7 Kosar, A., and Peles, Y., 2006, Convective Flow of Refrigerant R-123
Across a Bank of Micro Pin Fins, Int. J. Heat Mass Transfer, 49, pp. 3142
3155.
8 Krishnamurthy, S., and Peles, Y., 2007, Gas-Liquid Two-Phase Flow Across a
Bank of Micro Pillars, Phys. Fluids, 194, p. 043302.
9 Kosar, A., and Peles, Y., 2007, Boiling Heat Transfer in a Hydrofoil-Based
Micro Pin Fin Heat Sink, Int. J. Heat Mass Transfer, 5056, pp. 1018
1034.
10 Kosar, A., and Peles, Y., 2007, Micro Scale Pin Fin Heat Sinks Parametric
Performance Evaluation Study, IEEE Trans. Compon. Packag. Technol.,
304, pp. 855865.
11 Siu-Ho, A. M., Qu, W., and Pfefferkorn, F. E., 2006, Pressure Drop and Heat
Transfer in a Single-Phase Micro-Pin-Fin Heat Sink, ASME Paper No. IMECE2006 14777.
12 Krishnamurthy, S., and Peles, Y., 2008, Two-Phase Flow Pattern Transition
Across Micropillars Size Scale Effect at the Micro Scale, Phys. Fluids, 2, p.
023602.
13 Bjorg, R. W., Hall, G. R., and Rohsenow, W. M., 1982, Correlation of Forced
Convection Boiling Heat Transfer Data, Int. J. Heat Mass Transfer, 25, pp.
753757.
14 Bowring, R. W., 1962, Physical Model Based on Bubble Detachment and
Calculation of Steam Voidage in the Subcooled Region of a Heated Channel,
Institutt for Atomenergi, Halden, Norway, Paper No. Hpr-10.
15 Bergles, A. E., and Rohsenow, W. M., 1964, The Determination of Forced
Convection, Surface Boiling Heat Transfer, ASME J. Heat Transfer, 86, pp.
365372.
16 Kutateladze, S. S., 1961, Boiling Heat Transfer, Int. J. Heat Mass Transfer,
4, pp. 3145.
17 Kandlikar, S. G., 1998, Heat Transfer and Flow Characteristics in Partial
Boiling, Fully Developed Boiling, and Significant Void Flow Regions of Subcooled Flow Boiling, ASME J. Heat Transfer, 120, pp. 395401.
18 Rouhani, S. Z., and Axelsson, E., 1970, Calculation of Void Volume Fraction
in the Subcooled and Quality Boiling Regions, Int. J. Heat Mass Transfer, 13,
pp. 383393.
19 Lahey, R. T., 1978, A Mechanistic Subcooled Boiling Model, Proceedings of

041007-10 / Vol. 132, APRIL 2010

20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37

the Sixth International Heat Transfer Conference, Vol. 1, Toronto, Canada, pp.
292297.
Del Valle, V. H. M., and Kenning, D. B. R., 1985, Subcooled Flow Boiling at
High Heat Flux, Int. J. Heat Mass Transfer, 28, pp. 19071920.
Basu, N., Warrier, G. R., and Dhir, V. K., 2005, Wall Heat Flux Partitioning
During Subcooled Flow BoilingPart I: Model Development, ASME J. Heat
Transfer, 127, pp. 131140.
Basu, N., Warrier, G. R., and Dhir, V. K., 2005, Wall Heat Flux Partitioning
During Subcooled Flow BoilingPart II: Model Validation, ASME J. Heat
Transfer, 127, pp. 141148.
Warrier, G. R., and Dhir, V. K., 2006, Heat Transfer and Wall Heat Flux
Partitioning During Subcooled Flow Nucleate Boiling, A Review, ASME J.
Heat Transfer, 128, pp. 12431256.
Huang, L.-D., and Witte, L., 2001, Highly Subcooled Boiling in Crossflow,
ASME J. Heat Transfer, 123, pp. 10801085.
Huang, H. L., and Witte, L. C., 1996, An Experimental Investigation of the
Effects of Subcooling and Velocity on Boiling of Freon-113, ASME J. Heat
Transfer, 118, pp. 436441.
Chen, J. C., 1966, Correlation for Boiling Heat Transfer to Saturated Fluids in
Convective Flow, I&EC Process Des. Dev., 53, pp. 322329.
Shah, M. M., 2005, Improved General Correlation for Subcooled Boiling
Heat Transfer During Flow Across Tubes and Tube Bundles, HVAC&R Res.,
112, pp. 285303.
Cornwell, K., 1990, The Influence of Bubbly Flow on Boiling From a Tube in
a Bundle, Int. J. Heat Mass Transfer, 3312, pp. 25792584.
Gupta, A., 2005, Enhancement of Boiling Heat Transfer in a 5 3 Tube
Bundle, Int. J. Heat Mass Transfer, 48, pp. 37633772.
Lee, P. C., Tseng, F. G., and Pan, C., 2004, Bubble Dynamics in Micro
Channels. Part I: Single Microchannel, Int. J. Heat Mass Transfer, 47, pp.
55755589.
Kuo, C.-J., Kosar, A., Peles, Y., Virost, S., Mishra, C., and Jensen, M., 2006,
Bubble Dynamics During Boiling in Enhanced Surface Microchannels, J.
Microelectromech. Syst., 156, pp. 15141527.
Lie, Y., and Lin, T., 2006, Subcooled Flow Boiling Heat Transfer and Associated Bubble Characteristics of R-134a in a Narrow Annular Duct, Int. J.
Heat Mass Transfer, 49, pp. 20772089.
Martn-Callizo, C., Palm, B., and Owhaib, W., 2007, Subcooled Flow Boiling
of R-134a in Vertical Channels of Small Diameter, Int. J. Multiphase Flow,
33, pp. 822832.
Han, C., and Griffith, P., 1965, The Mechanism of Heat Transfer in Nucleate
Pool Boiling, Bubble Initiation, Growth and Departure, Int. J. Heat Mass
Transfer, 8, p. 887904.
Mikic, B. B., and Rohsenow, W. M., 1969, A New Correlation of PoolBoiling Data Including the Effect of Heating Surface Characteristics, ASME
J. Heat Transfer, 4414, pp. 245250.
Krishnamurthy, S., and Peles, Y., 2010, Flow Boiling on Micropin Fins Entrenched Inside a MicrochannelFlow Patterns and Bubble Departure Diameter and Bubble Frequency, ASME J. Heat Transfer, 1324, p. 041002.
Donnelly, B., ODonovan, T. S., and Murray, D. B., 2009, Surface Heat
Transfer Due to Sliding Bubble Motion, Appl. Therm. Eng., 297, pp. 1319
1326.

Transactions of the ASME

Downloaded From: http://asmedigitalcollection.asme.org/ on 04/18/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use